Brassinosteroid and gibberellin coordinate rice seed germination and embryo growth by regulating glutelin mobilization

Min Xiong,Lingyi Chu,Qinfeng Li,b,*,Jiwen Yu,Yiho Yng,Peng Zhou,Yong Zhou,b,Chngqun Zhng,b,Xiolei Fn,b,Dongsheng Zho,b,Chngjie Yn,b,Qioqun Liu,b,*

a Jiangsu Key Laboratory of Crop Genetics and Physiology/Jiangsu Key Laboratory of Crop Genomics and Molecular Breeding/Sate Key Laboratory of Hybrid Rice,College of Agriculture,Yangzhou University,Yangzhou 225009,Jiangsu,China

b Co-Innovation Center for Modern Production Technology of Grain Crops of Jiangsu Province/Key Laboratory of Plant Functional Genomics of the Ministry of Education,Yangzhou University,Yangzhou 225009,Jiangsu,China

Keywords:Seed germination Glutelin mobilization Brassinosteroid Gibberellin Oryza sativa L.

ABSTRACT Seed germination is the beginning of a new lifecycle,and involves many complex physiological and biochemical reactions including seed reserve mobilization in the endosperm and nutrient transport and reuse in the embryo.Although glutelin is a dominant storage protein in rice,its contribution to seed germination and its regulatory mechanisms are mostly unknown.Gibberellin(GA)and brassinosteroid(BR),two major growth-promoting phytohormones,also play positive roles in controlling seed germination.However,how GA and BR interact and coordinate seed germination and facilitate glutelin mobilization remains unclear.In the present study,biochemical and physiological analyses of seed germination indicated that both GA and BR promote seed germination and post-germination growth.Exogenous application of GA restored germination defects caused by BR deficiency or insensitivity.Proteomic and qRT-PCR results showed that the expression of several glutelin proteins and their encoding genes was induced by BR and GA in the embryo.Expression assays suggested that the increased accumulation of glutelin protein in the embryo was due to the accelerated degradation of glutelin by a cysteine proteinase(REP-1)in the endosperm.The breakdown of glutelin in the endosperm showed a strict positive correspondence with the length of the shoot.The GluA2 mutation led to reduced degradation rate of glutelin and defects in seed germination,and the promotion effect of GA on seed germination was weakened in the glua2 mutant.In vitro culture assay of rice embryos showed that glutelin mobilization functioned downstream of the GA and BR pathways to promote shoot elongation.These findings suggest a mechanism that mediates crosstalk between BR and GA in co-regulating rice seed germination and embryo growth.

1.Introduction

Seed storage proteins (SSPs) are major components (around 10%) of cereal kernels.According to their solubilities,SSPs can be divided into four categories:acid/alkaline-soluble glutelins,water-soluble albumins,alcohol-soluble prolamins,and salinesoluble globulins [1,2].Rice,as a staple food,provides more than one-fifth of the calories consumed by people worldwide.Interestingly,the composition of rice SSPs is different from that of other cereal counterparts,including wheat,maize,and barley.Glutelins are the dominant SSPs in rice,accounting for about 80%,whereas those in other cereals are prolamins [3,4].Owing to differences in composition,rice has a higher digestibility and absorption rate than other cereal crops [5].Glutelin not only is a determinant of rice nutritional quality,but also affects rice eating and cooking qualities [6].

Rice seeds contain glutelins of three different molecular weights,including a 57 kDa precursor glutelin,a 37–39 kDa acidic(α) subunit,and a 20–22 kDa basic (β) subunit.Based on the similarity of amino acid sequences,the glutelin-encoding genes can be classified into four categories:GluA,GluB,GluC,andGluD[7].In rice endosperm,glutelins are synthesized in the form of a pro-glutelin precursor and accumulated in protein body II(PB-II).The precursor protein is finally processed into mature 37-kDa acidic and 20-kDa basic subunits in the middle and late stages,respectively,of seed development [8–10].

Seed germination,the first step of the life cycle of rice,is critical in ensuring high yield.Generally,germination is a series of ordered physiological and morphological changes following the onset of swelling,including mobilization of storage reserves,repair of DNA,protein synthesis,cell wall loosening,and other processes.It typically begins with rapid uptake of water and ends with radicle protrusion [11].As germination progresses,seed reserves are degraded gradually,providing energy and metabolites for germination and seedling establishment[12].Thus,efficient degradation and utilization of seed reserves,including starch and SSPs,are particularly important during seed germination and post-germination growth.

Starch is the most abundant reserve in rice kernels and its degradation during germination is understood.It occurs by the secretion of gibberellins (GAs) from the embryo to the endosperm cells through the scutellum,triggering the early mobilization of storage reserves in the aleurone layer [13].GA in aleurone cells promotes the synthesis of amylase,which is then secreted to promote the degradation of starch to glucose [14–17].This process is reported [18,19] to be directly correspondence to the germination rate.In addition to its role in promoting amylase biosynthesis and starch degradation,GA induces the synthesis of proteinases that trigger the degradation of SSPs in seeds [12,20,21].

Given that proteins are polymers with low dispersibility in a dissolved state,they are difficult to pass through the cell wall.Accordingly,the SSP must be hydrolyzed into amino acids that are transported to the growth site of the embryo,where they recombine to form new proteins.Generally,the degradation of plant SSPs can be divided into three steps:modification of SSP,its degradation into peptides by endopeptidase,and further degradation of peptides into amino acids by peptidases such as dipeptidase,tripeptidase,and carboxypeptidase [22].Glutelin,the major SSP in rice,is readily digested by protease to provide amino acids,nitrogen,and sulfur during germination[23].In rice,the major protease reported to digest glutelin is a cysteine proteinase (REP-1),whose expression is also induced by GA and antagonized by abscisic acid (ABA);REP-1 thus represents a key regulator of glutelin degradation [24,25].

Although the regulation of seed germination has been extensively studied,the molecular mechanisms by which this occurs remain largely unknown.Generally,GA and ABA are the two major hormones antagonistically regulating seed germination.Other hormones,including brassinosteroid(BR),also play important roles in controlling seed germination.BR,a growth-promoting hormone,plays diverse roles in regulating plant growth and development,including cell elongation,seed germination,microtubule arrangement,cell division,and differentiation.BR interacts with both GA and ABA to coordinate seed germination [26,27].For instance,BRs rescued the germination of a GA-insensitive mutantsly1-2inArabidopsis[28].TheMFT(Mother of FLOWERING LOCUS T and TERMINAL FLOWER 1) family of genes was reported [29] to regulate seed germination via the BR signaling pathway inArabidopsis.Owing to a negative feedback regulatory mechanism between theMFTgene and ABA,BR promotes seed germination by abolishing the effect of ABA on seed dormancy[30].GA promotes the expression of the ABA catabolism geneABA8′OH-1in the aleurone layer,indicating a direct interaction between GA and ABA [31].These results suggest that GA,ABA and BR reciprocally modulate one another and thus form a complex regulatory network in coordinating seed germination.However,the mechanisms by which these phytohormones modulate seed germination by regulating the mobilization of seed reserves,in particular SSPs,are largely unknown.

Promoting seed germination and identifying its intrinsic regulatory mechanisms are focuses of seed research.High germination is necessary for high yields of rice,especially under direct-seeding cultivation conditions.Both BR and GA,two major growthpromoting phytohormones,promote seed germination.Our previous proteomics studies [27,32,33] have identified proteins regulated by both BR and GA during seed germination.The present study was aimed at determining how BR and GA coordinate the mobilization of SSPs in rice.

2.Materials and methods

2.1.Plant materials

Four different rice materials (Oryza sativassp.japonica) were used:Nipponbare (Nip),a mildly BR-insensitive mutantd61[34],a BR-deficient mutantgns4and its correspondingGNS4-RNAi lines[35],aglua2mutant and its chromosomal segment substitution line(CSSL) SL431 [6],andd61,a mutant of the BR receptor encoding geneBRI1,provided by Prof.Makoto Matsuoka(Nagoya University,Japan).SL431 was selected from a CSSL population derived from a cross of anindicacultivar,Habataki,and ajaponicacultivar,Sasanishiki,in our previous study [6].gns4,a novel allele ofDWARF11(D11),was identified in the ethyl methanesulfonate(EMS)-treated japonica cultivar Zhonghua 11C.The wild-type recipient for generation ofGNS4-RNAi lines was Nip.All rice plants were grown under the same climatic and management conditions during the summer months in a paddy field at Yangzhou University,Yangzhou,China.Mature seeds from superior rice spikelets were collected for germination analysis.

2.2.Seed germination analysis

For each experiment,100 dehulled rice seeds were sterilized with 70% ethanol and washed twice with Milli-Q water.Sterilized seeds were placed in 10×10 cm culture plates for germination and imbibed in solutions with four chemical treatments:2 μmol L-1brassinazole (BRZ,a specific BR biosynthetic inhibitor),10 μmol L-1GA,70 μmol L-1cycloheximide(CHX,a protein synthesis inhibitor) or the same amount of the solvent dimethylsulfoxide(DMSO;control).The seeds were then germinated in darkness in an artificial climate incubator at a temperature of 26 °C and a relative humidity of 70%.Seeds with radicle longer than 1 mm were considered as successfully germinated [36].Germination rates and seed weight were recorded every 12 h.The lengths of shoots of germinated seeds were measured at the indicated time points.For expression analysis,the embryo and endosperm were peeled off using a surgical blade,quick-frozen in liquid nitrogen,and held in a freezer at -80 °C.Each seed germination assay included at least three independent biological replicates,and each replicate contained 30 germinated seeds.

2.3.RNA extraction and quantitative real-time PCR analysis

Total RNA was extracted from rice tissues with a RNAprep pure Plant Kit (Tiangen,Beijing,China).At least 40 embryos and 20 endosperms were used for RNA extraction.About 1 μg of total RNA pre-treated with DNase I was used for reverse transcription in the SuperScript first strand cDNA synthesis system (Invitrogen,Van Allen Way Carlsbad,CA,USA).qRT-PCR was then performed on a real-time PCR system (ABI7500,Foster City,CA,USA) with SYBR Premix Ex Taq.Ubiquitin conjugase(OsUBC) was selected as the reference gene for normalization.Three biological replicates were included for each sample,and the primer details for qRTPCR are presented in Table 1.

2.4.Total protein extraction from seeds and SDS-PAGE analysis

A detailed total protein extraction method has been previously described [8].Briefly,embryo or endosperm tissues were ground into fine powder in liquid nitrogen.Total seed protein extraction buffer (4% SDS,4 mol L-1urea,5% 2-mercaptoethanol,and 125 mmol L-1Tris-HCl pH 6.8)was added to the powder to extract total proteins.After thorough mixing,the mixture was incubated at 37°C with shaking for 3 h and the suspension was then centrifuged at 4 °C for 20 min.The supernatant was aspirated into a new centrifuge tube and stored at 4 °C until further use.

An equal amount of total protein sample was mixed thoroughly with 2× sample buffer,and denatured at 99 °C for 10 min.Next,the sample was separated via SDS-PAGE using a 12% separation gel and 5% stacking gel.Following electrophoresis,the separation gel was stained with Coomassie Brilliant Blue G-250 and destained with 10%acetic acid.The protein amounts of the samples were calculated and compared using ImageJ software.

2.5.In vitro culture of embryos

Dehulled seeds were sterilized in 70% (v/v) ethanol for 10 min and then rinsed twice with sterilized water.The plumules were carefully separated from the embryos of seeds 2 h after imbibition(HAI)and transferred to a new culture dish forin vitroculture[37].Different concentrations(0,3,6,30 μg mL-1)of bovine serum albumin (BSA) or extracted rice glutelin proteins were added to the Milli-Q water for culturing the isolated rice embryos.After five days of growth under these conditions,the lengths of elongated shoots originating in the plumules were measured.The shoot samples were then quick-frozen in liquid nitrogen and stored at -80 °C for subsequent expression analysis.

2.6.Extraction and concentration determination of glutelin

Ground powder (100 mg) from endosperm tissues of germinated seeds was added to a centrifuge tube with 1 mL Milli-Q water.To remove albumin,the mixture was shaken for 1.5 h,followed by centrifugation at 10,000×gfor 15 min,and the supernatant discarded.Globulin and prolamin were similarly removed with 5%NaCl and 70%ethanol,respectively.After the above procedure was repeated three times,1 mL of 0.1 mol L-1NaOH solution was added to the residual precipitate and shaken for 1.5 h followed by centrifugation at 10,000×gfor 15 min and collection of the supernatant.The concentration of the glutelin solution was determined with a BCA Protein Assay Kit (Tiangen).

2.7.Generation of glua2 mutant

The target site ofGlutelin type-A2(OsGluA2)for gene editing was designed with a CRISPR/Cas9 system [33] and the target sequence was then introduced into vector pC1300-Cas9 (Table 1).The successfully generated construct was introduced into CSSL line SL431 viaAgrobacterium-mediated transformation [6].After sequencing and screening,homozygousglua2mutants were selected.

2.8.Data processing and analysis

In this report,most statistics are presented as means±standard deviation(SD).For experiments with single pairwise comparisons,Student’st-test was used to identify significant differences(*,P<0.05,**,P<0.01).For experiments with multiple comparisons,means were compared with Duncan’s multiple range test atP<0.05.

3.Results

3.1.Reciprocal effects of BR and GA on regulation of seed germination and embryo growth

The germination rate and fresh weight increase of BRZ-treated Nip seeds were markedly lower than those of the control (Fig.1a and b).The shoots of germinated seeds 96 h after imbibition(HAI)were much shorter than those of controls(Figs.1c,S1).To further evaluate BR effect on seed germination,BR-deficient mutantgns4andGNS4-RNAi transgenic lines were used for analysis.A single nucleotide deletion in the promoter ofGNS4led to smaller kernels and reduced expression ofGNS4(Fig.S2a–d).Both the germination rate and shoot lengths of thegns4mutant were lower than those of the wild-type control (Fig.S2e and f).Similar results were observed inGNS4-RNAi transgenic lines (Fig.S3a–f).Thus,deficiency in BR biosynthesis repressed seed germination.Correspondingly,the expression of some genes whose products promote cell elongation was down-regulated in embryos treated with BRZ,includingBRASSINOSTEROID-INSENSITIVE 1(OsBRI1) [38],Gibberellins 3β-hydroxylase 2(OsGA3ox2) [39],High tillering and dwarf2(HTD2)[40],Small and round seed 3(SRS3)[41],andBrittle culm1 like 4(OsBC1L4) [42] (Fig.S4).These results suggested that BR might regulate germination by modulating the expression of these genes.

Co-application of GA with BRZ not only restored seed germination,but also resulted in longer shoots than in controls(Figs.1,S1).A similar result was observed whend61,a BR-insensitive mutant,was treated with GA,showing increased germination rate and shoot length(Fig.S5a–c).These results suggested that GA compensated the suppressive effect of BR deficiency or insensitivity on seed germination.

3.2.GA compensates germination deficiency of BR-deficient or BRinsensitive seeds by up-regulating glutelin expression in embryos

An isobaric tags for relative and absolute quantification(iTRAQ)proteomic study was performed to identify proteins co-regulated by both GA and BR in embryos from germinated seeds,36 HAI.Comparison of proteins differentially expressed in embryos treated with BRZ alone or together with GA revealed four glutelin proteins:GluA1,GluA2,GluA3,and OsEnS-32 (GluB2).GA treatment markedly increased the accumulation of all four glutelins under BRdeficient conditions (Fig.2a).qRT-PCR results showed that the expression of all four genes was down-regulated by BRZ treatment.GA stimulated the expression ofGluA1,GluA2,andGluA3,but suppressed the expression ofOsEnS-32(Fig.2b).The expression of several other glutelin-encoding genes,includingGluB1,GluB5,GluBX,GluC,GluD1andGlu7,was consistent with the qRT-PCR findings:BRZ down-regulated and GA up-regulated the transcription of all detected glutelin genes(Fig.2c).In general,the expression profiles of glutelins were consistent at both transcript and protein levels in response to BRZ and GA treatment.Thus,promotion by GA of glutelin expression could be a mechanism by which GA rescues the seed germination deficiency of BR mutants.

3.3.Mobilization of glutelin in endosperm corresponds to shoot elongation

Fig.1.GA rescues seed germination defects of BR-deficient rice.(a)Germination rate of Nipponbare(Nip)seeds in response to hormonal treatments during the germination process.BRZ,BR biosynthesis inhibitor;GA,GA3;DMSO,solvent control.Fresh weight increase of 30 seeds (b),and shoot length of the germinated seeds at 96 h after imbibition(c).Each data point represents the mean of at least three independent experiments with standard deviation(SD).DMSO,dimethyl sulfoxide;BRZ,brassinazole;GA,GA3.*, P <0.05;**, P <0.01 (Student’s t-test).

Fig.2.Expression of glutelin is regulated by BR and GA in the embryo.(a) Under BR-deficient conditions,GA promoted glutelin accumulation in rice embryos,based on quantitative proteomic results.(b)Validation of proteomic data at the transcript level by qRT-PCR.(c)Measuring the transcription of other glutelin genes by qRT-PCR.Total protein and RNA were extracted from the embryos of germinated seeds 36 h after imbibition.OsUBC(Ubiquitin-conjugating enzyme)was used as the internal reference gene in qRT-PCR analysis.Error bars represent SDs.Each experiment was repeated at least three times.DMSO,dimethyl sulfoxide;BRZ,brassinazole;GA,GA3.Glu,Glutelin;OsEnS-32,Endosperm-specific Gene 32.*, P <0.05;**, P <0.01 (Student’s t-test).

Although the embryo,containing most of the genetic information of rice,plays a decisive role during seed germination,the supply of carbon,nitrogen,phosphorus,and minerals from hydrolysis events in the endosperm are also essential to successful germination.Glutelin,the dominant protein in the endosperm,is degraded into amino acids by proteolytic enzymes,and then transported to the embryo to support plumule and radicle elongation.To investigate how phytohormones modulate glutelin degradation in the endosperm,SDS-PAGE was performed to monitor the amount of glutelin accumulation at the beginning of germination (2 HAI),and after hormone treatment for four days (96 HAI).The residual glutelin in the endosperm was markedly decreased,either with or without hormone treatment 96 HAI (Fig.3a),suggesting that glutelin in the endosperm is mobilized and degraded during seed germination.The amount of residual glutelin in BRZ-treated seeds was greater than that of DMSO-treated controls.Similar results were observed in thegns4mutant andGNS4-RNAi transgenic lines:the amount of residual glutelin was greater in either of these than in its wild type 96 HAI (Fig.S6a and b).These results showed that BR deficiency indeed suppressed glutelin degradation.In addition,GA fully counteracted the negative effect of BRZ and greatly reduced the residual glutelin content (Fig.3a).The degradation rate of glutelin in the endosperm was speculated to be positively corresponding to seed germination.Considering that shoot length is corresponding to seed germination (Fig.1d),we suggest that shoot length could be an indicator of endosperm glutelin degradation and seed germination.

Fig.3.Glutelin degradation in the endosperm during seed germination.(a) Residual glutelin amounts in endosperm of germinated seeds treated with BRZ and/or GA.(b)Expression of REP-1 in embryo and endosperm of germinated seeds two days after imbibition.Error bars represent SDs.Each experiment was repeated at least three times.DMSO,dimethyl sulfoxide;BRZ,brassinazole;GA,GA3.M,protein marker.**,P<0.01 (Student’s t-test).

Next,the expression ofREP-1,encoding a key cysteine proteinase associated with the degradation of glutelin in rice,was measured to determine whether GA-and BR-promoted glutelin accumulation corresponded to glutelin metabolism.qRT-PCR analysis indicated that the expression ofREP-1was inhibited by BRZ but that this inhibition could be rescued by GA in the embryo(Fig.3b).To investigate the source–sink communication between endosperm and embryo during seed germination,the expression ofREP-1in the endosperm was evaluated.Overall,the hormoneregulated expression pattern ofREP-1in the endosperm was consistent with that in the embryo.However,the transcript abundance ofREP-1in the endosperm was more than 100 times higher than that in the embryo(Fig.3b).This finding suggests that hormone-regulated glutelin degradation potentially affects the seed germination process by controlling energy and resource supply for embryo growth.These results suggest that both mechanisms:phytohormone-promoted degradation of glutelin in the endosperm and increased synthesis of glutelin in embryos,could be major factors in the accumulation of glutelin in the embryo and could accelerate seed germination.

3.4.Promoting effect of BR and GA on germination and postgermination growth rely mainly on modulating glutelin degradation

To determine whether the regulation of glutelin degradation by BR and GA during germination was dependent on newly synthesized proteins,CHX was used to block protein biosynthesis.No marked difference in shoot lengths was observed between different groups of germinating seeds 60 HAI,including groups treated with DMSO,BRZ,BRZ+GA,and GA(Fig.4a),while the glutelin content of the last two groups supplemented with GA was less than the DMSO or BRZ treated samples (Fig.4b).Each group of germinated seeds was divided into two subgroups,one of which was further treated with CHX and the other treated as a mock control.After treatment for 36 h,the subgroup treated with CHX germinated more slowly than the control subgroup,indicating that new protein synthesis was essential for normal post-germination growth.However,the difference in shoot lengths among different phytohormone treatments was similar between the two subgroups;BRZ alone slightly suppressed the shoot length,while supplementing with GA strongly promoted shoot growth (Fig.4a).In addition,residual glutelin in the endosperm was more abundant in the CHX-treated subgroup,and glutelin degradation was also positively corresponding to the shoot lengths (Fig.4b).

The above results suggested that the hormonal regulation of proteolytic enzyme biosynthesis started at the early stages of seed germination,leading to subsequent differential shoot growth.CHX treatment blocked all protein synthesis indiscriminately,including that of proteolytic enzymes,such as REP-1,without affecting the differences among the various hormone-treated groups (Fig.4a and b).The above results also further verified a positive relationship between the shoot lengths and the degradation rate of glutelin in the endosperm.

3.5.BR and GA promote shoot elongation partially via the glutelin pathway

To eliminate potential interference from the endosperm and verify the direct correlation between glutelin degradation and shoot growth,anin vitrogerm culture system was used.After the seeds were imbibed for 2 h,the intact embryos were isolated and cultured in Petri dishes supplemented with different concentrations of glutelin solutions.In general,glutelin-supplied plumules were longer than the control at 120 HAI,and the promoting effect of glutelin on plumules was concentrationdependent.The lowest concentration of glutelin (3 μg mL-1)showed the most significant effect(Fig.5a and b).When the same concentrations of BSA were used as control to treat the isolated embryos,no significant effect was observed on embryo growth(Fig.S7a).To identify the mechanism by which glutelin stimulated plumule elongation in this system,several candidate genes with roles in stimulating cell elongation were analyzed by qRT-PCR.The expression ofGA-insensitive dwarf 1(GID1) andCarotenoid cleavage dioxygenase 7(CCD7) was increased by glutelin treatment(Fig.5c),while the expression ofHTD2andSRS3was not changed(Fig.S7b).Next,BRZ and GA treatments were also introduced into thein vitroculture system with or without the presence of 3 μg mL-1glutelin.Generally,glutelin promoted shoot elongation independent of BR and GA.In contrast,BRZ suppressed,and further treatment with GA recovered shoot elongation in the culture system,independent of glutelin(Fig.5d and e).These findings suggest the presence of other molecular mechanisms in GA-and BR-regulated seed germination and post-germination growth,in addition to glutelin.

Fig.4.Glutelin degradation in endosperm of germinated seeds requires new protein biosynthesis.(a) Shoot length and (b) glutelin accumulation of germinated seeds in response to treatments with phytohormone and cycloheximide(CHX).Nip seeds germinated for 60 h under treatment with DMSO,BRZ and/or GA were then divided into two groups,one group with CHX and the other with mock control for 36 h.Residual glutelin amounts in the endosperm were evaluated by SDS-PAGE before or after CHX treatment.Error bars represent SDs.Each experiment was repeated at least three times.DMSO,dimethyl sulfoxide;BRZ,brassinazole;GA,GA3.Error bars represent SDs.**,P<0.01 (Student’s t-test).

To verify that glutelin is critical for seed germination,aglua2mutant was used for germination analysis.This is a mutant with a 1-bp deletion(C)near the protospacer-adjacent motif(PAM)site ofGluA2leading to premature termination of translation (Fig.6a).SDS-PAGE assay showed that the amount of glutelin inglua2seeds was decreased by about 20% in the initial stage of germination(Fig.S8a).The germination rate ofglua2was lower than that of the wild type (Fig.6b).The shoots ofglua2germinated seeds at 96 HAI were shorter than those of the wild type(Fig.6c,S8b).Glutelin mobilization analysis indicated that the degradation rate of glutelin was slower inglua2than that in the wild-type (Fig.S8a),further confirming our hypothesis that the degradation rate of glutelin in the endosperm was positively associated with seed germination.Next,we treated theglua2mutant and its wild type with GA to confirm whether the germination-promotion effect of GA was weakened whenGluA2was knocked out.The results showed that although the germination rate and shoot lengths of bothglua2mutant and its wild type were increased by GA,the promotion effect of GA was attenuated in theglua2mutant(Fig.6b and c).This finding suggested that glutelin plays an important role in GApromoted rice germination.

4.Discussion

4.1.Seed reserve mobilization is essential for normal germination

Source–sink communication and its regulation by phytohormones during seed germination is an important topic of plant science research.Starch and storage proteins are two major components of rice seeds,accounting for about 90% and 10% respectively,of total dry weight of endosperm.Based on the nutritional model of germination,there is a supply of carbon flowing from the endosperm (source) to the embryo and its derivatives (sink)[43].At the source,the conversion of starch to glucose is catalyzed by hydrolytic enzymes,which include mainly α-amylase and betaamylase [44].Under oxygen-deficient conditions,starch is hydrolyzed by α-amylase to generate sugar substrates and then enters glycolysis to produce ATP [45].At the sink,the carbon source required by heterotrophic embryos mostly travels through the scutellum in the form of sucrose [46].Thus,with sufficient sugar,transient starch is synthesized and accumulated in the vascular tissue region of the embryo [37].The α-amylase promoter is then activated to combat sugar starvation[47].When the sugar content in embryos increases,the expression ofRAmy3D,which provides energy for shoot elongation [48],is inhibited.This suppressive effect of sugar is due to the differential sensitivity of various amylases to sugar.For example,RAmy1Ais more suitable for a highsugar environment thanRAmy3DandRAmy3E[17].In addition to the abovementioned starch degradation processes,ATP release,sucrose transport,and sugar utilization also affect germination.

Fig.5. In vitro plumule elongation assay in response to hormone and glutelin treatment.(a) Physiological morphology and (b) quantitative data of Nip plumule length following treatment with various concentrations of glutelin for 120 h.(c)Expression of GA-insensitive dwarf 1(GID1)and Carotenoid cleavage dioxygenase 7(CCD7)in response to glutelin treatment.(d)Physiological morphology and(e)quantitative data of Nip plumule length in response to hormone and glutelin treatments for 120 h.A concentration of 3 μg mL-1 glutelin was used in(d)and(e).Error bars represent SDs.Each experiment was repeated at least three times.DMSO,dimethyl sulfoxide;BRZ,brassinazole;GA,GA3.Values with different letters represent significant differences by Duncan’s multiple range test at P <0.05.

Fig.6.Germination analysis of glua2 mutant.(a)Schematic diagram of the target in OsGluA2 gene.The target site is underlined in yellow and the PAM(protospacer-adjacent motif)is marked with the green line.The deletion is indicated by the red hyphen.The blue letter indicates the new amino acid sequence after the single nucleotide deletion and the asterisk indicates the early termination of translation.Germination rate(b)and shoot length(c)of germinated seeds at 96 HAI with the treatment of GA or its mock DMSO.DMSO,dimethyl sulfoxide;GA,GA3.Error bars represent SDs.Each experiment was repeated at least three times.WT,wild-type.*, P <0.05;**, P <0.01 (Student’s ttest).Values with different letters represent significant differences by Duncan’s multiple-range test at P <0.05.

Besides starch mobilization,protein in the endosperm,which acts as a nitrogen source,also plays a critical role in germination.Once the seed begins to absorb water,the stored protease may trigger the degradation of a small amount of stored protein at a specific site,implying that the seed releases the mechanism by which the storage proteins are not degraded during the maturation period.This ensures a constant flow of nitrogen from the endosperm to the embryo and embryonic axes(Fig.7).To maintain normal amino acid supply,plants degrade proteins more rapidly under nutritional stress [49].For example,more REP-1 can be synthesized to degrade glutelin in rice endosperm under nitrogen starvation conditions [50].

Previous reports on seed reserve mobilization and germination focused mainly on starch,and there have been few studies of the relationship between glutelin metabolism and germination.In the present study,glutelin decreased gradually in the endosperm as germination progressed(Fig.3a),indicating that glutelin breakdown is closely corresponding to seed germination and shoot elongation (Figs.3,4).Glutelin treatmentin vitropromoted plumule elongation.The germination defect of theglua2mutant was due largely to the reduced degradation rate of glutelin(Figs.6,S8).Previous report[51]indicated that down-regulating the expression ofOsGluA2,OsEnS-94,andOsGlb1strikingly delayed rice seed germination.Interestingly,in the present study,higher concentrations of glutelin treatment did not result in longer shoots (Fig.5a and b),suggesting the presence of a feedback regulatory mechanism,such as sugar suppression.Similarly in buckwheat,cysteine proteinase,controlling the mobilization of stored 13S globulin,was inhibited by excessive degradation products of proteins [52,53].Identification of the mechanism by which glutelin metabolism mediates source–sink communication between endosperm and embryo during germination awaits further studies.

Fig.7.A simplified model illustrating hormonal regulation of glutelin mobilization during rice seed germination.BR and GA can promote the expression of cysteine proteinase(REP-1)in both the embryo and endosperm of seeds during germination.REP-1 in the endosperm increases the degradation of the stored glutelins,thus providing more amino acids for the growth of embryos.The small molecular amino acids generated in the endosperm are then transported to the embryo through the scutellum.Some of the amino acids are used as the direct nitrogen source for embryo growth,some are synthesized as transient glutelin for future embryo growth use,and the others are used for synthesizing structural or functional proteins.Then the expression of REP-1 induced by GA and BR in the embryos further stimulates the breakdown of the stored transient glutelins for promoting shoot elongation.

4.2.BR is involved in multi-hormonal crosstalk in regulation of seed germination and embryo growth

Phytohormones not only orchestrate intrinsic developmental programs,but also convey environmental signals to regulate plant growth and development as well as adaption to the environment[54].Multiple phytohormones are integrated into a regulatory network to coordinate various plant growth and development events.BR is also involved in modulation of seed germination;however,the molecular mechanism by which this occurs is unclear.The rice dwarf mutantd61germinates more slowly than the wild-type control,indicating that the BR signaling pathway plays a positive role in seed germination [34].In this study,both BRZ treatment and knockout/down BR biosynthesis genes suppressed seed germination to some extent,confirming the promoting effect of BR on seed germination(Figs.1a–c,S2,S3).BR strongly induced the expression of the GA biosynthesis geneOsGA3ox2[55],a finding consistent with the reduced expression ofOsGA3ox2by BRZ treatment in germinated embryo (Fig.S4).InArabidopsis,BRASSINOSTEROID INSENSITIVE2 BIN2 (BIN2) interacted with the transcription factor ABA INSENSITIVE 5 (ABI5) to positively regulate response to ABA during seed germination [56].This result is consistent with the finding[28]that BR inhibited ABA signaling.Additive and synergistic effects of BR and GA and synergy between BR and auxin have also been observed[57].In present study,GA antagonizes the suppressive effect of BRZ on germination rate,fresh weight increase,and shoot elongation.Similar results were observed in the BRinsensitive mutantd61(Fig.S5).We suggest that the promotion of seed germination by BR must rely on a functional GA pathway.

4.3.Modulating glutelin mobilization is one mechanism mediating GAand BR-coordinated regulation of seed germination and embryo growth

The proteomics and qRT-PCR results of this study indicate that co-regulation of storage protein metabolism by GA and BR represents a novel mechanism mediating their interaction in promoting seed germination and post-germination growth (Fig.7).Our SDS-PAGE results showed that BR and GA promoted glutelin degradation in the endosperm and thus resulted in less glutelin accumulation(Fig.3a,S6).One possible explanation is that the expression of a key cysteine proteinase,REP-1,responsible for glutelin degradation,was induced in the endosperm by BR and GA(Fig.3b).Generally,more glutelin consumption in the endosperm will stimulate the transfer of more amino acids into the embryo,thus promoting germination.It is possible that the increased amino acids in embryos (caused by hormone-induced storage protein mobilization in the endosperm)are temporarily stored in the form of glutelin for the subsequent elongation of shoots and roots.The structural and functional proteins that maintain the life activities of embryos are also synthesized from these amino acids (Fig.7).

iTRAQ proteomic quantitative analysis identified the target proteins involved in the synergistic regulation of seed germination by BR and GA,with several glutelins involved (Fig.2a).qRT-PCR results further confirmed the regulation of expression of glutelin genes by GA and BR during seed germination (Fig.2b and c).Inin vitroculture experiments of plumules,BR and GA increased shoot elongation,while glutelin further increased their promoting effects(Fig.5).Knockout ofGluA2not only inhibited seed germination,but also weakened the promotion by GA of germination,indicating that glutelin is required for GA-promoted germination(Fig.6b and c).A simplified model(Fig.7)is proposed to illustrate the hormonal regulation of glutelin mobilization to promote rice seed germination.Briefly,BR and GA induce the expression ofREP-1during seed germination.REP-1 in the endosperm degrades the storage protein glutelin,providing amino acids for the growth of the embryos.The amino acids generated in the endosperm are then transported to the embryo through the scutellum for germination and post-germination growth.The accumulated amino acids in the embryo are used as the direct nitrogen source for embryo growth,as transient glutelin for future embryo growth,or as substrates for synthesizing structural and functional proteins.

5.Conclusions

The present study suggests a novel mechanism by which GA and BRcould coordinate seed germination from the perspective of storage protein mobilization.A positive correspondence was established between glutelin consumption in the endosperm and germination events in the embryo.We hypothesize that storage protein mobilization is a universal regulatory mechanism explaining not only BR-GA crosstalk,but also other phytohormone interactions in co-regulating seed germination and embryo growth.

CRediT authorship contribution statement

Qianfeng Li and Qiaoquan Liu:conceived and designed the research.Min Xiong,Lingyi Chu,Jiawen Yu,Yihao Yang,Peng Zhou,and Changquan Zhang:performed the experiments.Min Xiong,Qianfeng Li,Yong Zhou,Changjie Yan,Xiaolei Fan,and Dongsheng Zhao:analyzed the data,Min Xiong and Qianfeng Li:wrote the manuscript.All authors discussed the results and commented on the manuscript.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

This work was financially supported by the National Natural Science Foundation of China (31825019),Science Fund for Distinguished Young Scholars of Jiangsu Province (BK20200045),Postgraduate Research &Practice Innovation Program of Jiangsu Province (KYCX18_2369),Jiangsu Six Talent Peaks (SWYY-154),Jiangsu PAPD,‘‘333”,and Qinglan,Innovative and Entrepreneurial Talent Project.

Appendix A.Supplementary data

Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2020.11.006.